From an individual to a collective dimension of data protection

Transcript

computer law & security review 32 (2016) 238–255
Available online at www.sciencedirect.com
ScienceDirect
w w w. c o m p s e c o n l i n e . c o m / p u b l i c a t i o n s / p r o d c l a w. h t m
Personal data for decisional purposes in the age
of analytics: From an individual to a collective
dimension of data protection
Alessandro Mantelero a,b,*
a
Nanjing University of Information Science & Technology (NUIST), School of Public Administration, Jiangsu,
China
b
Polytechnic University of Turin, Department of Management and Production Engineering, and Nexa Center for
Internet & Society, Turin, Italy
A B S T R A C T
Keywords:
In the big data era, new technologies and powerful analytics make it possible to collect and
Big data
analyse large amounts of data in order to identify patterns in the behaviour of groups, com-
Right to privacy
munities and even entire countries.
Data protection
Existing case law and regulations are inadequate to address the potential risks and issues
Group privacy
related to this change of paradigm in social investigation. This is due to the fact that both
Collective interests
the right to privacy and the more recent right to data protection are protected as indi-
Data protection authorities
vidual rights. The social dimension of these rights has been taken into account by courts
Risk assessment
and policymakers in various countries. Nevertheless, the rights holder has always been the
data subject and the rights related to informational privacy have mainly been exercised by
individuals.
This atomistic approach shows its limits in the existing context of mass predictive
analysis, where the larger scale of data processing and the deeper analysis of information
make it necessary to consider another layer, which is different from individual rights. This
new layer is represented by the collective dimension of data protection, which protects
groups of persons from the potential harms of discriminatory and invasive forms of data
processing.
On the basis of the distinction between individual, group and collective dimensions of
privacy and data protection, the author outlines the main elements that characterise the
collective dimension of these rights and the representation of the underlying interests.
© 2016 Alessandro Mantelero, Published by Elsevier Ltd. All rights reserved.
* Department of Management and Production Engineering, Politecnico di Torino, C.so Duca degli Abruzzi, 24, 10120 Torino, Italy.
E-mail address: [email protected].
http://dx.doi.org/10.1016/j.clsr.2016.01.014
0267-3649/© 2016 Alessandro Mantelero, Published by Elsevier Ltd. All rights reserved.
1.
Introduction and scope of the analysis
Big data analytics make it possible to infer predictive information from large amounts of data in order to acquire further
knowledge about individuals and groups, which may not necessarily be related to the initial purposes of data collection.1
Moreover, analytics group people together by their qualitative attributes and habits (e.g. low-income people, “workingclass mom”, “metro parents”2) and predict the future behaviour
of these clusters3 of individuals.4
This approach is adopted, for instance, by some health insurance companies, which extract predictive information about
the risks associated with segments of clients on the basis of
their primetime television viewing, propensity to buy general
merchandise, ethnicity, geography or use of mail order buying.5
In these cases, predictions based on correlations6 do not only
affect individuals, which may act differently from the rest of
the group to which have been assigned,7 but also affect the
whole group and set it apart from the rest of society. An
example in this sense is provided by the “neighbourhood’s
1
See David Bollier, ‘The Promise and Perils of Big Data’ (Aspen
Institute, Communications and Society Program 2010) <http://
www.aspeninstitute.org/sites/default/files/content/docs/pubs/
The_Promise_and_Peril_of_Big_Data.pdf> accessed 27 February 2014.
See also Pertti Ahonen, ‘Institutionalizing Big Data methods in social
and political research’ (2015) Big Data & Society 1–12 <http://
bds.sagepub.com/content/2/2/2053951715591224> accessed 21 July
2015.
2
This is one of the categories used by US data brokers to define
specific segments of population based on models of predictive
behaviour. In this sense, the category “metro parents” includes consumers “primarily in high school or vocationally educated [. . .]
handling single parenthood and the stresses of urban life on a small
budget”, see Federal Trade Commission, ‘Data Brokers: A Call for
Transparency and Accountability’ (2014), 20 and Appendix B <https://
www.ftc.gov/system/files/documents/reports/data-brokers-call
-transparency-accountability-report-federal-trade-commission
-may-2014/140527databrokerreport.pdf> accessed 27 February 2014.
3
In this article, the notion of cluster is used to identify a set of
individuals that are directly or indirectly grouped on the basis of
common qualitative elements (class of age, habits, geographic distribution, etc.).
4
See Recital nn. 51, 58 and 58a of the Proposal for a Regulation
of the European Parliament and of the Council on the protection
of individuals with regard to the processing of personal data and
on the free movement of such data (General Data Protection Regulation) text adopted by the Council of the European Union, Brussels,
19 December 2014 (hereinafter abbreviated as EU Proposal).
5
See Satish Garla, Albert Hopping, Rick Monaco and Sarah
Rittman, ‘What Do Your Consumer Habits Say About Your Health?
Using Third-Party Data to Predict Individual Health Risk and Costs.
Proceedings’ (SAS Global Forum 2013) <http://support.sas.com/
resources/papers/proceedings13/170-2013.pdf> accessed 28 February
2015; see also Federal Trade Commission (n 2) 20 and Appendix B.
6
See Bollier (n 1). See also Mireille Hildebrandt, ‘Profiling: From
Data to Knowledge. The challenges of a crucial technology’ (2006)
30(9) Datenschutz und Datensicherheit 548.
7
See also Barbara D. Underwood, ‘Law and the Crystal Ball: Predicting Behavior with Statistical Inference and Individualized
Judgment’ (1979) 88 Yale Law Journal 1408.
239
general credit score” adopted by credit companies,8 which
induces companies to provide opportunities for people living
in a given neighbourhood in a way that bears no relationship
to their individual conditions, but is based on the aggregate
score of the area.9
These issues are not new and may be considered the effect
of the evolution of profiling technologies, in a context
characterised by an increased volume of information available and powerful software analytics.10 Nevertheless, previous
forms of categorisation and profiling were based on a few standard variables (e.g. sex, age, family income, marital status, place
of residence); therefore, their predictive ability was limited.
Today, big data analytics use hundreds of different variables
to infer predictive information about groups of people and, in
many cases, these variables concern aspects that are not clearly
related to the final profiles created by analytics.
Moreover, users are often unaware of these forms of data
analysis and of the impact that some information may have
on their membership of one or another group created by analytics. Finally, decision makers use the outcomes generated by
big data analytics to take decisions that affect individuals and
groups, without allowing them any participation in the process,
which remains primarily based on obscure data management and frequently takes place in situations of imbalance
between data gatherers and data subjects.
In the light of the above, the use of big data analytics creates
“a new truth regime”,11 in which general strategies are adopted
8
This score predicts credit risks of individuals that live in a small
geographic area and it is defined on the basis of aggregate credit
scores.
9
See Pam Dixon and Robert Gellman, ‘The Scoring of America:
How Secret Consumer Scores Threaten Your Privacy and Your Future’
(2014), 21, 44, <http://www.worldprivacyforum.org/wp-content/
uploads/2014/04/WPF_Scoring_of_America_April2014_fs.pdf>
accessed 10 March 2015. See also Frank Pasquale, The Black Box Society.
The Secret Algorithms That Control Money and Information (Harvard University Press 2015) 22–26; Danielle Keats Citron and Frank Pasquale,
‘The Scored Society: Due Process For Automated Predictions’ (2014)
89 Wash. L. Rev. 1; Meike Kamp, Barbara Körffer and Martin Meints,
‘Profiling of Customers and Consumers – Customer Loyalty Programmes and Scoring Practices’ in Mireille Hildebrandt and Serge
Gutwirth (eds.), Profiling the European Citizen. Cross-Disciplinary Perspective (Springer 2010) 205–211; Anton H. Vedder, ‘Privatization,
Information Technology and Privacy: Reconsidering the Social Responsibilities of Private Organizations’ in Geoff Moore (ed), Business
Ethics: Principles and Practice (Business Education Publishers 1997)
215–226.
10
See also Serge Gutwirth and Mireille Hildebrandt, ‘Some Caveats
on Profiling’ in Serge Gutwirth, Yves Poullet and Paul de Hert (eds.)
Data protection in a profiled world (Dordrecht, London 2010) 32–33.
11
See Antoinette Rouvroy, ‘Des données sans personne: le
fétichisme de la donnée à caractère personnel à l’épreuve de
l’idéologie des Big Data’ (2014) 9 <http://works.bepress.com/
antoinette_rouvroy/55> accessed 8 March 2015; Antoinette Rouvroy,
‘Algorithmic Governmentality and the End(s) of Critique’ (2013)
<https://vimeo.com/79880601> accessed 10 March 2015.
240
on a large scale on the basis of representations of society
generated by algorithms, which predict future collective
behaviour.12 These strategies are then applied to specific individuals, given the fact that they are part of one or more groups
generated by analytics.13
The use of analytics and the adoption of decisions based
on group behaviour rather than on individuals are not limited
to commercial and market contexts. They also affect other important fields, such as security and social policies, where a
different balancing test should be applied, given the importance of public interest issues.
12
See Pasquale (n 9); Viktor Mayer-Schönberger and Kenneth
Cukier, Big Data. A Revolution That Will Transform How We Live, Work
and Think (John Murray 2013); Bollier (n 1); McKinsey Global Institute Big data: The next frontier for innovation, competition, and productivity
(2011) <http://www.mckinsey.com> accessed 16 April 2012. See also
Bellagio Big Data Workshop Participants, ‘Big data and positive social
change in the developing world: A white paper for practitioners
and researchers’ (Oxford Internet Institute 2014) <http://
www.rockefellerfoundation.org/uploads/files/c220f1f3-2e9a-4fc6
-be6c-45d42849b897-big-data-and.pdf> accessed 28 June 2015; Ira
S. Rubinstein, ‘Big Data: The End of Privacy or a New Beginning?’
(2013) 3 (2) International Data Privacy Law 74–87; danah boyd and
Kate Crawford, ‘Six Provocations for Big Data’ (paper presented at
Oxford Internet Institute’s “A Decade in Internet Time: Symposium on the Dynamics of the Internet and Society”, Oxford,
September 21, 2011) <http://ssrn.com/abstract=1926431> accessed 16 April 2012; danah boyd and Kate Crawford, ‘Critical
Questions for Big Data: Provocations for a Cultural, Technological, and Scholarly Phenomenon’ (2012) 15(5) Information,
Communication, & Society 662–679; Omer Tene and Jules Polonetsky,
‘Privacy in the Age of Big Data. A Time for Big Decisions’ (2012) 64
Stan. L. Rev. Online 63–69 <http://www.stanfordlawreview.org/sites/
default/files/online/topics/64-SLRO-63_1.pdf> accessed 14 March
2013.
13
See Federal Trade Commission (n 2) IV–V (“Potentially sensitive categories include those that primarily focus on ethnicity and
income levels, such as ‘Urban Scramble’ and ‘Mobile Mixers,’ both
of which include a high concentration of Latinos and African Americans with low incomes. Other potentially sensitive categories
highlight a consumer’s age such as ‘Rural Everlasting,’ which includes single men and women over the age of 66 with ‘low
educational attainment and low net worths,’ while ‘Married Sophisticates’ includes thirty-something couples in the ‘uppermiddle class . . . with no children’ ”). See also Bollier (n 1);
Hildebrandt, ‘Profiling: From Data to Knowledge. The challenges of
a crucial technology’ (n 6) 549–550.
One example of this is provided by predictive policing solutions like “PredPol”,14 a software used by US local police forces
to anticipate, prevent and respond more effectively to crime,
on the basis of cross check data, places and techniques of recent
crimes. PredPol and similar software are able to predict future
crimes and their location, but they also induce a “self-fulfilling
cycles of bias”. This is due to the fact that police departments allocate more resources to the areas suggested by
analytics and this increases crime detection at local level, with
the result of reinforcing the original prediction. At the same
time, a reduced police presence in other areas reduces crime
detection and produces an adverse prediction for these areas.15
The consequence of these software solutions is a potential geographical discrimination, which might not directly affect
individuals, but has an impact on local communities in terms
of social stigma or inadequate provision of police services. In
this sense, there is a collective interest in a correct and accurate use of data.16
These scenarios show the collective dimension of decisions adopted using data analytics and their potential bias.17
Against this background, Korzybski’s statement “a map is not
the territory”18 sums up the focus of this article. The logic of
the author of the map, the way in which the territory is
14
See Walter L. Perry, Brian McInnis, Carter C. Price, Susan C. Smith
and John S. Hollywood, ‘Predictive Policing. The Role of Crime Forecasting in Law Enforcement Operations’ (The RAND Corporation
2013), <http://www.rand.org/content/dam/rand/pubs/research
_reports/RR200/RR233/RAND_RR233.pdf> accessed 10 March 2015.
See also Pasquale (n 9) 41–46, 48–51; Robinson + Yu, ‘Civil Rights,
Big Data, and Our Algorithmic Future. A September 2014 report on
social justice and technology’ (2014), 18–19 <http://bigdata.fairness.io/
wp-content/uploads/2014/09/Civil_Rights_Big_Data_and_Our
_Algorithmic-Future_2014-09-12.pdf> accessed 10 March 2015;
Alessandro Mantelero and Giuseppe Vaciago, ‘Social media and big
data’ in Babak Akhgar, Andrew Staniforth and Francesca Bosco (eds.)
Cyber Crime & Cyber Terrorism. Investigator’s Handbook (Elsevier 2014),
175–196; Andrew Guthrie Ferguson, ‘Predictive Policing: The Future
of Reasonable Suspicion’ (2012) 62 Emory L. J. 259 <http://
www.law.emory.edu/fileadmin/journals/elj/62/62.2/Ferguson.pdf>
accessed 29 January 2014; Rosamunde van Brakel and Paul De Hert,
‘Policing, surveillance and law in a pre-crime society: Understanding the consequences of technology based strategies’ (2011) 20(3)
Journal of Police Studies 20(3): 163 – 192 http://www.vub.ac.be/
LSTS/pub/Dehert/378.pdf accessed 27 July 2015.
15
See Kelly K. Koss, ‘Leveraging Predictive Policing Algorithms to
Restore Fourth Amendment Protections in High-Crime Areas in a
Post-Wardlow World’ (2015) 90 Chi.-Kent. L. Rev. 301, 311–312.
16
See also Perry et. al. (n 14), 118–125; Oscar H. Gandy Jr., ‘Exploring Identity and Identification in Cyberspace’ (2000) 14 Notre Dame
J.L. Ethics & Pub. Pol’y 1085, 1100 <http://scholarship.law.nd.edu/
ndjlepp/vol14/iss2/10> accessed 10 July 2015.
17
See also Tal Z. Zarsky, ‘Transparent Predictions’ (2013) 4 U. Ill.
L. Rev. 1503, 1510–1513.
18
Alfred Korzybski, ‘A Non-Aristotelian System and its Necessity for Rigour in Mathematics and Physics’ in Alfred Korzybski (ed)
Science and sanity: An Introduction to Non-Aristotelian Systems and General
Semantics (Institute of General Semantics 1933) 747, 750 <http://
lipn.univ-paris13.fr/~duchamp/Books&more/Neurosciences/
Korzybski/%5BAlfred_Korzybski%5D_Science_and_Sanity_An
_Introduc%28BookFi.org%29.pdf> accessed 15 March 2015; see also
Kate Crawford, ‘Algorithmic Illusions: Hidden Biases of Big
Data’, presentation at Strata 2013, <https://www.youtube.com/
watch?v=irP5RCdpilc> accessed 15 March 2015.
represented and the potential errors of representation can
produce different maps of the same territory. Maps are not
neutral. In the same way, in social investigations, the strategies used to group data, the logic of big data analytics and their
potential bias have an influence on the final representation of
groups and society.19
This “categorical” approach characterising the use of analytics leads policymakers to adopt common solutions for
individuals belonging to the same cluster generated by analytics. These decisional processes do not consider individuals
per se, but as a part of a group of people characterised by some
common qualitative factors.
This leads to a reflection on privacy and data protection.20
The use of personal information and big data analytics to
support decisions exceeds the boundaries of the individual dimension and assumes a collective dimension, with potential
harmful consequences for some groups.21 In this sense, prejudice can result not only from the well-known privacy-related
risks (e.g. illegitimate use of personal information, data security), but also from discriminatory and invasive forms of data
processing.22
The dichotomy between individuals and groups is not new
and it has already been analysed with regard to the legal aspects
of personal information. Nonetheless, the right to privacy and
19
It should be noted that different architectures of algorithms may
produce different results, although on the basis of the same factors.
See Dixon and Gellman (n 9), 2 (“new consumer scores use thousands of pieces of information about consumers’ pasts to predict
how they will behave in the future. Issues of secrecy, fairness of
underlying factors, use of consumer information such as race and
ethnicity in predictive scores, accuracy, and the uptake in both use
and ubiquity of these scores are key areas of focus”).
20
The origin of data, the nature of information and its legal protection (i.e. right to privacy or data protection) are not particularly
relevant in the context described in the previous paragraphs. In
this article, the analysis of privacy and data protection is more
focused on the use of information and on the relationship between
individual and collective dimensions, rather than on the traditional aspects of secrecy and data quality. See also Fred H. Cate
and Viktor Mayer-Schönberger, ‘Data Use and Impact. Global
Workshop’ (The Center for Information Policy Research and The
Center for Applied Cybersecurity Research, Indiana University 2013)
iii <http://cacr.iu.edu/sites/cacr.iu.edu/files/Use_Workshop
_Report.pdf> accessed 27 February 2014; Alessandro Mantelero, ‘The
future of consumer data protection in the E.U. Rethinking the “notice
and consent” paradigm in the new era of predictive analytics’ in
this Review (2014), vol 30, issue 6, 643–660.
21
See also Kate Crawford, Gustavo Faleiros, Amy Luers, Patrick
Meier, Claudia Perlich and Jer Thorp, ‘Big Data, Communities and
Ethical Resilience: A Framework for Action’ (2013) 6–7 <http://
www.rockefellerfoundation.org/app/uploads/71b4c457-cdb7-47ec
-81a9-a617c956e6af.pdf> accessed 5 April 2015; danah boyd, Karen
Levy, Alice Marwick, ‘The Networked Nature of Algorithmic Discrimination’, in Seeta Peña Gangadharan, Virginia Eubanks and Solon
Barocas Data and Discrimination: Collective Essays (Open
Technology Institute and New America 2014) 56 <http://www
.newamerica.org/downloads/OTI-Data-an-Discrimination-FINAL
-small.pdf> accessed 14 April 2015.
22
See also The White House, Executive Office of the President, ‘Big
Data: Seizing Opportunities, Preserving Values’ (2014) <http://
www.whitehouse.gov/sites/default/files/docs/big_data_privacy
_report_may_1_2014.pdf> accessed 26 December 2014. See also
Zarsky (n 17) 1560–1563; Vedder (n 9).
241
the (more recent) right to the protection of personal data have
been largely safeguarded as individual rights, despite the social
dimension of their rationale.23
The focus on the model of individual rights is probably the
main reason for the few contributions by privacy scholars on
the collective dimension of privacy and data protection. Hitherto, only few authors have investigated the notion of group
privacy. They have represented this form of privacy as the
privacy of the facts and ideas expressed by the members of a
group in the group environment or in terms of protection of
information about a group.
Against this background, this article is not an attempt to
provide a new interpretation of group privacy or to investigate the relationships between the legal and sociological notions
of the group, which is only briefly touched on in the following paragraphs. Rather, it focuses on the new kind of groups
that results from the use of big data analytics to represent the
“territory” of our society. In this light, the article investigates
the consequences of this algorithmic representation, in terms
of protection of collective rights.
From this perspective, the first part of this article deals with
the traditional notions of individual privacy and group privacy;
it points out the novelty of the groups generated by algorithms,24
which are ontologically different from the groups referred to
by the original notion of “group privacy”.25 In this sense, big
data analytics generate new groups, which did not previously exist in society, variable aggregations of individuals whose
personal information is mined in order to extract predictive
inferences.
The different origin and morphology of these groups make
it necessary to investigate the collective dimension of privacy
and data protection, which is different from the manifestation of the individual right to be let alone in the group context
or the protection of information regarding the group. For this
reason, the second part of the article focuses on the main elements that characterise this collective dimension in the
context of big data analytics26 and examines the nature of the
collective interests at issue in this regard, their representation and the balance with other conflicting interests.
2.
The notion of group in privacy literature:
Group privacy and individual rights
Privacy scholars have devoted few contributions to group privacy
and collective interests in data processing. A first approach considers group privacy as the right to privacy concerning
information shared within a group by its members.27 In this
sense, there is no autonomous notion of group privacy, but only
a peculiar attitude of individual privacy in the context of groups.
Individual privacy describes the conditions under which a “right
23
See below fn. 61.
Section 2.
25
See Edward J. Bloustein, ‘Group Privacy: The Right to Huddle’
(1977) 8 Rutgers-Cam. L.J. 219 and Edward J. Bloustein, Individual and
Group Privacy (Transaction Books 1978) 123–186; Alan F. Westin, Privacy
and Freedom (Atheneum 1970) 42–51.
26
Section 3.
27
See Bloustein, Individual and Group Privacy (n 25).
24
242
to be let alone” should be recognised, while group privacy determines the type of personal information sharing that goes
within a group.28 Group privacy therefore refers to the privacy
of the facts or ideas expressed by members of a group in the
group environment and provides a guarantee that this information will not be revealed outside the group.29
This notion of group privacy focuses on secrecy and intimacy and, for this reason, is mainly based on the level of trust
existing among the members of a group. The consequence is
a duty of confidentiality.30 Group privacy concerns the breach
of this duty.31 Nevertheless, this does not represent a change
in the traditional perspective, which remains based on the individual’s right to privacy.32
A slightly different notion of group privacy is represented
by the idea of “organizational privacy”,33 which focuses on
control over information concerning collective entities and on
the self-determination of these entities.34 In this sense, group
privacy still relies on confidentiality and regards the interests of the group itself in the protection of facts, acts or
decisions that concern its internal affairs and its organisational
autonomy.35 Thus questions regarding “organizational privacy”
do not only concern whether legal persons might have a legitimate claim to privacy,36 but also revolve around the indirect
protection of individuals who constitute the collective entities and their group interests.37
These two theories on group privacy concern the peculiar
nature of the sharing of personal information within a group.
They provide a sort of context-related notion of individual
privacy. A different approach focuses on information regarding groups per se and does not focus on groups as a sum total
of the individuals that make them up, with the related internal dynamics.38 In this perspective, a group is an autonomous
entity (an organised or a non-organised collective entity) and
28
See Bloustein, Individual and Group Privacy (n 25) 127–130.
See Bloustein, Individual and Group Privacy (n 25) 129–134. In the
description of the various contexts in which the right to privacy
is relevant in the light of the group dimension, the author considers marital, priest–penitent, lawyer–client and physician–
patient relationships. In all these cases, the right to privacy is mainly
related to intimacy and secrecy.
30
It should be noted that terms like confidentiality or “relational privacy” have been also used to describe the aspects
concerning the notion of group privacy that has been investigated by Bloustein. See, e. g., Lawrence O. Gostin, Public health law:
power, duty, restraint (University of California Press 2008) 316; Christine M. Emery, ‘Relational privacy. A Right To Grieve in The
Information Age: Halting The Digital Dissemination of DeathScene Images’ (2011) 42 Rutgers L. J. 765. Authors used the notion
of relational privacy to point out the contextual nature of this right,
its intimate nature and its protection, but without focusing on the
group dimension per se. See James Rachels, ‘Why Privacy is Important’ (1975) 4(4) Philosophy & Public Affairs 323–333; Charles Fried,
‘Privacy [A moral analysis]’ (1968) 77(3) Yale L. J. 475–493. See also
Kendall Thomas, ‘Beyond the Privacy Principle’ (1992) 92(6) Columbia Law Review 1431, 1445–1446. On the contrary, the study
conducted by Bloustein focuses on the group environment and provides a more detailed analysis of the right to privacy in this context.
Moreover, this author puts the notion of group privacy in relationship with the dynamics of groups and the sociological theories on
groups. Finally, the mentioned notion of “relational privacy” is very
vague, since it is used by legal scholars to describe different kinds
of social-related aspects concerning privacy, from privacy of the
relatives concerning the death of members of their family to
intimate sexual aspects, up to the more recent dimension of
social network interaction. See also Lorraine G. Kisselburgh,
‘Reconceptualizing privacy in technological realms: Theoretical
frameworks for communication’ (2008) Annual meeting of the International Communication Association, TBA, Montreal, Quebec,
Canada <http://citation.allacademic.com/meta/p233000_index.html>
accessed 20 February 2015; Beate Rössler, The value of privacy (Polity
2005) 130–133.
31
See Bloustein, Individual and Group Privacy (n 25) 137–140,
180–181.
32
See Bloustein, Individual and Group Privacy (n 25) 125 (“Group
privacy is an extension of individual privacy. The interest protected by group privacy is the desire and need of people to come
together, to exchange information, share feelings, make plans and
act in concert to attain their objectives”).
33
See Westin (n 25).
29
34
Collective entities may be autonomous and independent of the
sum of their members, but, at the same time, they are the sum of
the individual persons who make them up. For this reason, although organisational privacy can be considered as an autonomous
right of legal persons, in many cases it also represents an indirect protection of the individual rights of their members and of the
secrecy of members’ interaction in the context of the organisation.
See Westin (n 25) 42. See also Lee A. Bygrave, Data Protection Law.
Approaching Its Rationale, Logic and Limits (Kluwer Law International 2002) 175–176, 186.
35
See also Westin (n 25). For an analysis of the theoretical approach adopted by Westin, see also Bygrave, Data Protection Law.
Approaching Its Rationale, Logic and Limits (n 34) 247–252.
36
This issue is part of the more general debate on personality
rights of collective entities and on the dualism characterising legal
persons, from von Savigny and van Gierke to Kelsen, Hohfeld and
Hart. See Friedrich Karl von Savigny, Jural relations : or, The Roman
law of persons as subjects of jural relations : being a translation of the
second book of Savigny’s System of modern Roman law, translated by
W. H. Rattigan (Wildy & Sons 1884); Otto von Gierke, Die
Genossenschaftstheorie und die deutsche Rechtsprechung (Weidmann
1887); Wesley N. Hohfeld and Walter W. Cook, Fundamental legal conceptions as applied in judicial reasoning : and other legal essays (Yale
University Press 1923); Herbert L. A. Hart, Definition and theory in
jurisprudence (Stevens 1954); Hans Kelsen, Reine Rechtslehre. Mit einem
Anhang: Das Problem der Gerechtigkeit (F. Deuticke 1960). See also John
Dewey, ‘The Historic Background of Corporate Legal Personality’
(1926) 35(6) Yale L. J. 655–673; Katsuhito Iwai, ‘Persons, Things and
Corporations: The Corporate Personality Controversy and Comparative Corporate Governance’ (1999) 47(4) The American Journal
of Comparative Law 583–632.
37
On this dual dimension, which characterises collective entities and the legal protection of related interests, see Bygrave, Data
Protection Law. Approaching Its Rationale, Logic and Limits (n 34) 175–
176, 250–253.
38
See Bygrave, Data Protection Law. Approaching Its Rationale, Logic
and Limits (n 34) 173–298.
group privacy refers to information that identifies and describes the group.39
Under this third interpretation, group privacy protects information referring to collective entities – both legal persons
and organisations or groups without a formal and independent identity – and acts as an extension of individual data
protection to these entities.40 Although this notion of group
privacy is different from the other definitions briefly described above, it seems not to challenge the traditional
perspective that characterises privacy and data protection. The
group dimension affects the manifestation of these rights in
a specific context (the group), but they still revolve around the
model of individual rights, although referring to a collective
entity or its members.41 Nevertheless, this approach is the
closest to the collective dimension of data protection, mainly
when it focuses on non-organised collective entities.42
Despite these differences between the theories about group
privacy, this brief overview shows that the existing studies are
for the most part based on the individual rights model where
they consider the group dimension of privacy and data protection. These two rights are related to given individuals who
are members of a group, or to the group itself as an autonomous collective body. In both cases, the architecture of these
rights is not inspired by the idea of the group’s collective and
non-aggregative interests.
This approach to the issues related to groups is consistent with the traditional protection of the rights in question.
The right to privacy and the right to data protection have been
treated as individual rights in both the U.S. and European experiences, though based on differing origins and evolutions.
In the U.S., at the end of the 19th century, Warren and
Brandeis shaped the modern idea of privacy,43 which was different from the previous notion of protection of private life
39
and Limits (n 34) part III. The author extensively describes the various
issues concerning data protection rights of collective entities. See
also Jean-Pierre Chamoux, ‘Data Protection in Europe: The Problem
of the Physical Person and their Legal Person’ (1981) 2 J. Media Law
& Practice 70–83. See also Article 3(2)(b) of Convention for the Protection of Individuals with regard to Automatic Processing of
Personal Data (Council of Europe, Strasbourg, 28.01.1981) and Recital
24 of the Directive 95/46/EC.
40
and Limits (n 34) 186, 241–282, 288–289. See also Bart van der Sloot,
‘Do privacy and data protection rules apply to legal persons and
should they? A proposal for a two-tiered system’ in this Review
(2015), vol. 31, issue 1, 26, 33–45.
41
But see, more recently, Lee A. Bygrave and Dag Wiese Schartum,
‘Consent, Proportionality and Collective Power’ in Serge Gutwirth,
Yves Poullet, Paul De Hert, Cécile de Terwangne and Sjaak Nouwt
(eds.), Reinventing Data Protection? (Springer 2009), 157–173.
42
and Limits (n 34) 283–295. Bygrave points out the peculiar nature
of non-organised collective entities, which are created by persons
or organisations outside the group. Moreover, the author suggests some remedies to protect the interests of these entities and
their members according to the existing data protection framework.
43
Samuel D. Warren and Luis D. Brandeis, ‘The Right to Privacy’
(1890) 4(5) Harv. L. Rev. 193–220.
243
based on property.44 In spite of this, the right to privacy, although redefined as a personality right, remained largely based
on the individual dimension.45 Neither the notion of decisional privacy nor its constitutional dimension, originating in
the ground-breaking opinion given by Brandeis in his role as
Supreme Court judge,46 abandoned the individualistic nature
of the right.
On the other side of the Atlantic, individual privacy protection stemmed from the same social factors (the invasive
attitude of the “penny press” and new media) that justified the
response of the U.S. legal system to privacy invasion and the
protection of the right to be let alone.47 However, the European legal notion of privacy did not draw its origins from the
U.S. experience, but was independently shaped by legal scholars and the courts.48 From the theoretical point of view, the right
to privacy was placed in the sphere of individual rights, as in
the U.S., but in European case law and literature there is a closer
connection with the general theory of personality rights.49
Moreover, unlike in the U.S., the European notion of privacy
has not acquired the wider dimension of the U.S. decisional
privacy, but has remained more focused on informational
privacy. This does not mean the right to self-determination with
regard to government and public bodies has not been recognised
in Europe, but that it rests on the different fundamental
44
See Warren and Brandeis (n 43) 205, 213; Westin (n 25) 337–
345; David W. Leebron, ‘The Right to Privacy’s Place in the Intellectual
History of Tort Law’ (1991) 41 Case W. Res. L. Rev. 769, 775–778, 781;
Robert C. Post, ‘Rereading Warren and Brandeis: Privacy, Property
and Appropriation’ (1991) 41 Case W. Res. L. Rev. 647, 663–670; Amitai
Etzioni, The Limits of Privacy (Basic Books 1999) 189.
45
See Warren and Brandeis (n 43) 219; Westin (n 25) 330–364; Etzioni
(n 44) 189, 196; Daniel J. Solove, Understanding Privacy (Harvard University Press 2008) 12–37, 78–98.
46
See Brandeis’ opinions in Olmstead v. United States 277 US 438,
471 (1928). See also Sweezy v. New Hampshire 354 US 234 (1957);
NAACP v. Alabama 357 US 449 (1958); Massiah v. US 377 US 201 (1964);
Griswold v. Connecticut, 381 US 479 (1965); Roe v. Wade 410 US 113
(1973).
47
See Michael Schudson, Discovering the News. A Social History of
American Newspaper (Basic Books 1978) 12–60.
48
See, e.g., Trib. civ. Seine, 16 June 1858, D.P., 1858.3.62. See also
Lee A. Bygrave, ‘Privacy Protection in a Global Context. A Comparative Overview’ (2004) 7 Scandinavian Studies in Law 319, 326–
331; James Q. Whitman, ‘The Two Western Cultures of Privacy:
Dignity versus Liberty’ (2004) 113 Yale L.J. 1151–1221; Frits W. Hondius,
Emerging Data Protection in Europe (North Holland Publishing Company
1975) 6.
49
See Stig Stromhölm, Right of Privacy and Rights of Personality. A
comparative Survey (Norstedt & Soners 1967) 28–31. See also Hans
Giesker, Das Recht der Privaten an der eigenen Geheimsphäre. Ein Beitrag
zu der Lehre von den Individualrechten (Müller 1905); Josef Kohler,
Urheberrecht an Schriftwerken und Verlagsrecht (F. Enke 1907) 441.
244
freedoms recognised by European charters and conventions,
not solely on the right to privacy.50
Despite these differences, the nature of the right to privacy
depends primarily on the individual rights model on both sides
of the Atlantic.51 The collective dimension of this right has been
recognised in the U.S. and Europe, but protected mainly indirectly, as an aggregation of individual privacy issues and not
as an autonomous dimension.52
50
See the influential decision adopted by the Federal German Constitutional Court (Bundesverfassungsgericht), 15 December 1983,
Neue Juristische Wochenschrift, 1984. <https://www.zensus2011.de/
SharedDocs/Downloads/DE/Gesetze/Volkszaehlungsurteil_1983
.pdf?__blob=publicationFile&v=9> accessed 25 June 2014. For an
English translation, see Human Rights Law Journal 1984, 5: 94. See
also the article 8 of the European Convention on Human Rights
and the related case law of the European Court of Human Rights,
Council of Europe, ‘Case Law of the European Court of Human Rights
Concerning the Protection of Personal Data’ (2013) <http://
www.coe.int/t/dghl/standardsetting/dataprotection/TPD_documents/
DP%202013%20Case%20Law_Eng%20%28final%29.pdf> accessed 20
March 2015; European Court of Human Rights, ‘Protection of personal
data’ (2014) <http://www.echr.coe.int/Documents/FS_Data_ENG.pdf>
accessed 21 March 2015. See also Stefano Rodotà, ‘Data Protection
as a Fundamental Right’ in Gutwirth, Poullet, De Hert, de Terwangne
and Nouwt (n 41), 77–82; Joseph A. Cannataci, ‘Lex Personalitatis &
Technology-driven Law’ (2008) 5(1) SCRIPTed 1–6 <DOI: 10.2966/
scrip.050108.1>. On the notion of personality right, see Giorgio Resta,
‘Personnalité, persönlichkeit, personality’ (2014) 1(3) Comparative Perspectives on the Protection of Identity in Private Law 215–243.
51
See Bygrave, ‘Privacy Protection in a Global Context. A Comparative Overview’ (n 48) 324–325; Solove, Understanding Privacy (n
45) ch. 2. See also Colin J. Bennett and Charles D. Raab, The Governance of Privacy. Policy instruments in global perspective (Ashgate 2003)
ch. 1; Robert C. Post, ‘The Social Foundations of Privacy: Community and Self in the Common Law Tort’ (1989) 77 Cal. L. Rev. 957;
Julie E. Cohen, ‘Examined Lives: Informational Privacy and the
Subject’ (2000) 52 Stan. L. Rev. 1373, 1426–1428.
52
See inter alia Article 29 Data Protection Working Party, ‘Letter
to Mr. Larry Page, Chief Executive Officer’ (2013) <http://ec.europa.eu/
justice/data-protection/article-29/documentation/other-document/
files/2013/20130618_letter_to_google_glass_en.pdf> accessed 27
February 2014; Irish Data Protection Commissioner, ‘Facebook Ireland
Ltd. Report of Re-Audit’ (2012) <http://dataprotection.ie/documents/
press/Facebook_Ireland_Audit_Review_Report_21_Sept_2012.pdf>
accessed 27 February 2014; Italian Data Protection Authority, ‘Injunction and Order Issued Against Google Inc.’ (2013) <http://
www.garanteprivacy.it/web/guest/home/docweb/-/docweb-display/
docweb/3133945> accessed 27 February 2014. Only in a few
hypotheses, collective privacy is recognised as an autonomous dimension, which is different from individual privacy. This happens
in labour law, where the representatives of employees concur on the
adoption of the decisions concerning surveillance in the workplace on behalf of the workers. See, European Commission, ‘Second
stage consultation of social partners on the protection of workers’
personal data’ (undated) 7, 10, 16–17 <http://ec.europa.eu/social/
main.jsp?catId=708> accessed 10 January 2015. See also specific
references to the provisions of European national labour laws in Mark
Freedland, ‘Data Protection and Employment in the European Union.
An Analytical Study of the Law and Practice of Data Protection and
the Employment Relationship in the EU and Its Member’ (1999) 40–
43 <http://ec.europa.eu/social/main.jsp?catId=708> accessed 25
January 2015; Frank Hendrickx, ‘Protection of Workers’ Personal Data
in the European Union’ (undated) 33–35, 98–101, <http://ec.europa.eu/
social/main.jsp?catId=708> accessed 18 January 2015. See also Article
4 of the Italian labour statute (L. 300/1970). See below para 3.2.
The same considerations can be applied to the legal regimes
of personal information, which is regulated under data protection statutes. Although data protection laws have drawn their
origins from citizens’ concerns about government social
control,53 regarding the collective dimension of data protection, the regulation focuses on data subjects and their rights.54
Collective interests have been actively protected as the sum
total of various individual needs. Hence, lawmakers, the courts
and data protection authorities have addressed these interests with remedies that are mainly focused on individual rights
and their enforcement.
In light of the above, the approach based on the individual rights model, adopted by legal scholars with regard to
group privacy, is in line with the general legal notions of privacy
and data protection. It is also consistent with the theoretical
studies on group theory conducted in the field of sociology.
The various approaches of legal scholars seem to reflect the
more general controversy between individualistic and organic
sociological theories about the nature of groups.55 On the one
hand, attention to the individual dimension of privacy and the
interactions between different individuals56 is consistent with
the notion of group as the sum of the relationships existing
among its members (individualistic theory). On the other hand,
when the analysis takes into consideration information concerning the group itself as a whole,57 the group is seen as an
autonomous unit that assumes the form of an organised collective entity (organic theory).
In this context, the legal approach that considers group
privacy as relating to the individual’s privacy issues within a
group58 is in line with the individualistic theory, which sees
groups as entities in which individuals interact with each other
in a continuous and relatively stable manner. Moreover, from
a sociological perspective, the members of a group are aware
of being part of the group and the group is usually recognised
as an autonomous social structure. According to this position, a group is the product of concurrent decisions of various
persons who are striving to reach a common goal or share
common experiences, values or interests.
53
See Westin (n 25) 158–168, 298–326; Adam C. Breckenridge, The
Right to Privacy (University of Nebraska Press 1970) 1–3; Secretary’s Advisory Committee on Automated Personal Data Systems,
‘Records, Computers and the Rights of Citizens’ (1973) <http://
epic.org/privacy/hew1973report/> accessed 27 February 2014. See
also Solove, Understanding Privacy (n 45) 4–5. See also Myron Brenton,
The Privacy Invaders (Coward-McCann 1964); Vance Packard, The Naked
Society (David McKay 1964); Arthur R. Miller, The Assault on Privacy
Computers, Data Banks, Dossiers (University of Michigan Press 1971)
54–67; Viktor Mayer-Schönberger, ‘Generational development of data
protection in Europe?’ in Philip E. Agre and Marc Rotenberg (eds),
Technology and privacy: The new landscape (MIT Press 1997) 221–227.
54
See Colin J. Bennett, Regulating Privacy: Data Protection and Public
Policy in Europe and the United States (Cornell University Press 1992)
29–33, 47; Mayer-Schönberger, ‘Generational development of data
protection in Europe?’ (n 53) 219, 221–222. See also Bygrave and
Wiese Schartum (n 41) 169.
55
See Bloustein, Individual and Group Privacy (n 25) 124.
56
Ibid.
57
See Westin (n 25) 42; Bygrave, Data Protection Law. Approaching
Its Rationale, Logic and Limits (n 34) 173–282.
58
See Bloustein, Individual and Group Privacy (n 25) 124.
On the other hand, the organic theory of groups is consistent with the notion of group privacy in terms of “organizational
privacy”, 59 which focuses on the informational selfdetermination of the group itself. In this sense, group privacy
is more closely connected to the secrecy of the group’s activities than to the secrecy of personal information shared within
the group by its members. The organic theory is also consistent with the more recent approach focused on data
protection,60 which is not necessarily related to the secrecy of
information, but also regards publicly available data on collective entities.
Finally, as we have seen, the central role of the individual
rights model in protecting privacy and personal information
does not mean that legal systems disregard the social dimensions of these rights. Both privacy and data protection play an
important role in safeguarding not only individual interests,
but also the quality of society in general. Freedom of association, limits to disproportionate surveillance practices, and
prevention of discrimination based on sensitive personal data
are just few examples of the social effects of safeguarding the
right to privacy and personal information. Values such as democracy and pluralism are strictly related to the protection of
these rights.61
However, the courts may address issues related to general
interests only when they receive complaints from rights holders,
but the right holders may have no interest in these issues, be
unaware of the general interest,62 or be in no position to react
to potential threats to their interests, owing to situations of
power imbalance.63
Independent authorities may better address these issues of
general interest,64 but we should remember that these authorities often act on a discretionary basis and this may have
negative effects, in terms of under-deterrence. The licensing
model, which has been adopted in some cases by national regulators, represents a possible solution in assessing the risks
associated with specific technologies or business models, and
to prevent under-deterrence.65
For these reasons, independent authorities may play an important role in safeguarding interests related to the collective
dimension of privacy and data protection in the big data environment. Even so, adequate solutions are required to enlarge
59
See above fn. 33. See also Bygrave, Data Protection Law. Approaching Its Rationale, Logic and Limits (n 34) 250.
60
See above fn. 38. On the debate regarding the application of
privacy notion to collective entities, see Bygrave, Data Protection Law.
Approaching Its Rationale, Logic and Limits (n 34) 241–256.
61
See Simitis, ‘Auf dem Weg zu einem neuen Datenschutzrecht’
(1984) 3 Informatica e diritto 111; Schwartz, ‘Privacy and Participation: Personal Information and Public Sector Regulation in the
United States (1995) 80 Iowa L. Rev. 553, 560–561. For a general overview, see Bygrave, Data Protection Law. Approaching Its Rationale, Logic
and Limits (n 32) 133–143, 150–157; Anita L. Allen, Uneasy Access:
privacy for women in a free society (Rowman & Littlefield 1988) ch. 2.
62
Extensive video surveillance programs, which have been adopted
by municipalities or police departments, offer an example in this
sense: citizens are aware of being monitored, but, in many cases,
do not care about surveillance and are not interested in the social
impact of these control solutions.
63
See para 3.2.
64
See fn. 126.
65
See below para 3.3.
245
their role and move from a discretionary approach to a general
and mandatory assessment of the impact of technologies and
business models on data protection.66
3.
A new dimension of protection
In the big data era, new technologies and powerful analytics
make it possible to collect and analyse huge amounts of data
to try and identify patterns in the behaviour of groups of
individuals67 and to take decisions that affect the internal dynamics of groups, with consequences for the collective issues
of the people involved.
Nevertheless, these groups are different from those considered in the literature on group privacy, in fact that they are
created by data gatherers selecting specific clusters of information. Data gatherers shape the population they set out to
investigate. They collect information about various people, who
do not know the other members of the group and, in many
cases, are not aware of the consequences of their belonging
to a group.68 This is the case of consumer group profiling,69
scoring solutions70 and predictive policing applications,71 mentioned above.
The issues relating to privacy that arise from this new situation are different from the issues of individual privacy and
group privacy. We are neither in the presence of forms of analysis that involve only individuals, nor in the presence of groups
in the traditional sociological meaning of the term, given group
members’ lack of awareness of themselves as part of a group
and the lack of interactions among people grouped into various
clusters by data gatherers.
66
See below para 3.3.
Moreover, this is also possible without directly identifying data
subjects; see Andrej Zwitter, ‘Big Data ethics’ (2014) Big Data &
Society 1, 4–5. See also Paul Ohm, ‘Broken Promises of Privacy: Responding to the Surprising Failure of Anonymization’ (2010) 57 UCLA
L. Rev. 1701–1777; Philippe Golle, ‘Revisiting the uniqueness of simple
demographics in the US population’ in Ari Juels (ed), Proc. 5th ACM
workshop on Privacy in electronic society (ACM 2006) 77–80; Latanya
Sweeney, ‘Simple Demographics Often Identify People Uniquely’
(Carnegie Mellon University 2000) <http://dataprivacylab.org/
projects/identifiability/paper1.pdf> accessed 24 January 2015; Latanya
Sweeney, ‘Foundations of Privacy Protection from a Computer
Science Perspective’ in Proc. Joint Statistical Meeting, AAAS, Indianapolis (2000) http://dataprivacylab.org/projects/disclosurecontrol/
paper1.pdf, accessed 24 January 2015.
68
See Mireille Hildebrandt, ‘Defining Profiling: A New Type of
Knowledge?’ in Hildebrandt and Gutwirth (n 9) 19–20. See also Executive Office of the President of the United States-Council of
Economic Advisers, ‘Big Data Differential Pricing’ (2015) 18 <https://
www.whitehouse.gov/sites/default/files/docs/Big_Data_Report
_Nonembargo_v2.pdf> accessed 25 March 2015; Gandy (n 16) 1085,
1088, 1095; Hildebrandt, ‘Profiling: From Data to Knowledge. The
challenges of a crucial technology’ (n 6) 549–550.
69
See also Ryan Calo, ‘Digital Market Manipulation’ (2014) 82 George
Washington Law Review 995.
70
See Federal Trade Commission (n 2). But see Articles 18 and 20
of the Directive 2014/17/EU on credit agreements for consumers
relating to residential immovable property and amending
Directives 2008/48/EC and 2013/36/EU and Regulation (EU)
No 1093/2010.
71
See above fn. 14.
67
246
We must therefore extend the field of investigation to the
collective interests of the persons whose personal data are being
collected, analysed and grouped. The differing nature of these
groups of individuals requires a different approach that cannot
be exclusively based on individual rights.
The new scale entails the recognition of a new layer, represented by the rights of groups of individuals to the protection
of their collective privacy and data protection. Moreover, since
the predictive nature of big data analytics is designed to assist
decisions that affect a plurality of individuals in various fields,
we must also consider the social and ethical effects associated with this type of analysis.72
This kind of approach differs from the theoretical framework proposed by legal scholars in shaping the notion of
group privacy,73 but it can give a specific answer to the issues
arising from the present and future scenarios of the datadriven society.
3.1.
Collective data protection and its rationale
The collective dimension of data protection has its roots in the
individual’s right to privacy and shares some similarities with
group privacy, but differs from both these previous notions. On
the one hand, notions of individual privacy and data protection do influence the definition of the boundaries of this
collective dimension, but the greater scale affects the morphology of the interests involved and their enforcement. At the
same time, group privacy – as hitherto described by legal scholars – represents the notion that is closest to the idea of collective
data protection.
On the other hand, collective data protection does not necessarily concern facts or information referring to a specific
person,74 as with individual privacy and data protection. Nor
72
See Article 29 Data Protection Working Party, ‘Statement on
the role of a risk-based approach in data protection legal
frameworks’ (2014) 4 <http://ec.europa.eu/justice/data-protection/
article-29/documentation/opinion-recommendation/files/2014/
wp218_en.pdf> accessed 27 February 2014; Paul M. Schwartz,
‘Data Protection Law and the Ethical Use of Analytics’ 22–26
<http://www.huntonfiles.com/files/webupload/CIPL_Ethical
_Undperinnings_of_Analytics_Paper.pdf> accessed 27 February 2014;
David Wright, ‘A framework for the ethical impact assessment of
information technology’ (2011) 13 Ethics Inf. Technol. 199–226. See
also Luciano Floridi, The 4TH Revolution. How the Infosphere is Reshaping Human Reality (Oxford University Press 2014) 189–190; Helen
Nissenbaum, Privacy in Context. Technology, Policy, and the Integrity of
Social Life (Stanford University Press 2010) 231; Rayan M. Calo, ‘Consumer Subject Review Boards: A Thought Experiment’ (2013) 66 Stan.
L. Rev. Online 97, 101–102; Cynthia Dwork and Deirdre K. Mulligan, ‘It’s not Privacy and It’s not Fair’ (2013) 66 Stan. L. Rev. Online
35, 38; Bygrave, Data Protection Law. Approaching Its Rationale, Logic
and Limits (n 34) 61–62, 339; Julie E. Cohen, ‘What Privacy is For’ (2013)
126 Harv. L. Rev. 1904, 1925–1926; Crawford et al. (n 21) 4.
73
See above para 2.
74
In many cases, private companies and governments have no
interests in profiling single customers or citizens, but wish to discover the attitudes of clusters of individuals. Their main purpose
is to predict future behaviours of segments of the population to
achieve economic or political goals. See Bollier (n 1).
does it concern clusters of individuals that can be considered groups in the sociological sense of the term. In addition,
collective rights are not necessarily a large-scale representation
of individual rights and related issues.75 Finally, collective data
protection concerns non-aggregative collective interests,76 which
are not the mere sum of many individual interests.77
The importance of this collective dimension depends on the
fact that the approach to classification by modern algorithms does not merely focus on individuals, but on groups or
clusters of people with common characteristics (e.g. customer habits, lifestyle, online and offline behaviour, etc.).78 Data
gatherers are mainly interested in studying groups’ behaviour
and predicting this behaviour, rather than in profiling single
users. Data-driven decisions concern clusters of individuals and
only indirectly affect the members of these clusters. One
example of this is price discrimination based on age, habits
or wealth.
The most important concern in this context is the protection of groups from potential harm due to invasive and
discriminatory data processing. The collective dimension of data
processing is mainly focused on the use of information,79 rather
than on secrecy80 and data quality.
We need to adopt a broader notion of discrimination here,
one that encompasses two different meanings. In a negative
sense, discrimination is “the unjust or prejudicial treatment
of different categories of people”. In a more neutral and potentially positive sense, though, discrimination may be the
“recognition and understanding of the difference between one
thing and another”.81 Both these dimensions assume relevance in the context of big data analytics.
We will focus below on the first meaning, since the unfair
practices characterised by discriminatory purposes are generally
75
See Francesco Capotorti, ‘Are Minorities Entitled to Collective
International Rights?’ in Yoram Dinstein and Mala Tabory (eds), The
Protection of Minorities and Human Rights (Martinus Nijhoff Publishers 1992) 507, 511.
76
See Dwight G. Newman, ‘Collective Interests and Collective
Rights’ (2004) 49(1) American Journal of Jurisprudence 127, 131. See
also below in the present section. On the contrary, an aggregative
approach seems to be consistent with the notion of group privacy
described by Bloustein, Individual and Group Privacy (n 25) 123–186.
77
Contra Vedder (n 9), who claims that the notion of collective
privacy “reminds of collective rights”, but subjects of collective rights
are groups or communities. Conversely, the groups generated by
group profiling are not communities of individuals sharing similar
characteristics and structured or organised in some way. For this
reason, Vedder uses the different definition of “categorial privacy”,
see below fn. 101.
78
See above para 1 and below in the text.
79
See Cate and Mayer-Schönberger (n 20) iii; Mantelero, ‘The future
of consumer data protection in the E.U. Rethinking the “notice and
consent” paradigm in the new era of predictive analytics’ (n 20).
80
See Bloustein. Individual and Group Privacy (n 25) 182.
81
See <http://www.oxforddictionaries.com/it/definizione/inglese/
discrimination> accessed 29 January 2015.
247
forbidden and sanctioned by law.82 This article concerns involuntary forms of discrimination in cases where big data
analytics provide biased representations of society.83
For example, in 2013 a study examined the advertising provided by Google AdSense and found statistically significant racial
discrimination in advertisement delivery.84 Similarly, Kate Crawford has pointed out certain “algorithmic illusions”85 and
described the case of the City of Boston and its StreetBump
smartphone app to passively detect potholes. The application had a signal problem, due to the bias generated by the low
penetration of smartphones among lower income and older
residents. While the Boston administration took this bias into
account and solved the problem, less enlightened public officials might underestimate such considerations and make
potentially discriminatory decisions.86
Another example is the Progressive case, in which an insurance company obliged drivers to install a small monitoring
device in their cars in order to receive the company’s best rates.
The system considered as a negative factor driving late at night,
but did not take into account the potential bias against lowincome individuals, who are more likely to work night shifts,
compared with late-night party-goers, “forcing them [lowincome individuals] to carry more of the cost of intoxicated
and other irresponsible driving that happens disproportionately at night”.87
These cases represent situations in which a biased representation of groups and society results from flawed data
processing88 or a lack of accuracy in the representation.89 This
produces potentially discriminatory effects as a consequence
of the decisions taken on the basis of analytics.
On the other hand, the other sense of discrimination involving different treatment of different situations may represent
an intentional goal for policy makers, which is in line with the
rule of law. This is the case of law and enforcement bodies and
intelligence agencies, which adopt solutions to discriminate
between different individuals and identify targeted persons.
Here there is a deliberate intention to treat given individuals
differently, but this is not unfair or illegal providing it is within
existing legal provisions. Nonetheless, as in the previous case,
potential flaws or a lack of accuracy may cause harm to citizens.
For instance, criticisms have been raised with regard to the
aforementioned predictive software adopted in recent years
by various police departments in the US.90 Leaving aside the
constitutional profiles associated with these applications and
the peculiar balance of interests of this use of data, there have
been cases where people were named as potential offenders
due to merely remote connections with authors of serious
crimes.91 Criticisms also concern the use of risk assessment
procedures based on analytics coupled with a categorical approach (based on typology of crimes and offenders) in U.S.
criminal sentencing.92
82
See inter alia European Commission, ‘Developing AntiDiscrimination Law in Europe. The 28 EU Member States, the Former
Yugoslav Republic of Macedonia, Iceland, Liechtenstein, Norway and
Turkey compared’ (2013) <http://www.non-discrimination.net/
content/media/Developing%20Anti-Discrimination%20Law%20in
%20Europe%20EN%2029042014%20WEB.pdf> accessed 28 March
2015; Evelyn Ellis and Philippa Watson, EU Anti-Discrimination Law
(2nd edn Oxford University Press 2015). See also Article 14 of the
Convention for the Protection of Human Rights and Fundamental Freedoms; Article 21 of the Charter of Fundamental Rights of
the European Union; Article 19 of the Treaty on the Functioning
of the European Union; Directive 2000/43/EC; Directive 2000/78/
EC. See also Wim Schreurs, Mireille Hildebrandt, Els Kindt and
Michaël Vanfleteren, ‘Cogitas, Ergo Sum. The Role of Data Protection Law and Non-discrimination Law in Group Profiling in the
Private Sector’ in Hildebrandt and Gutwirth (n 9) 258–264.
83
See Citron and Pasquale (n 9), 14; Birny Burnbaum (2013). Insurers’ Use of Credit Scoring for Homeowners in Ohio: A Report
to the Ohio Civil Rights Commission.
84
See Latanya Sweeney, ‘Discrimination in Online Ad Delivery’
(2013) 56(5) Communications of the ACM 44–54. See also Bianca
Bosker, ‘Google’s Online Ad Results Guilty Of Racial Profiling, According To New Study’ The Huffington Post (2 May 2013) <http://
www.huffingtonpost.com/2013/02/05/online-racial-profiling_n
_2622556.html> accessed 27 March 2015.
85
Crawford (n 18).
86
See Kate Crawford, ‘The Hidden Biases in Big Data’ (2013) Harv.
Bus. Rev. April 1, 2013, <https://hbr.org/2013/04/the-hidden-biasesin-big-data> accessed 29 January 2015. Similar considerations can
be made in the case of the predictive policing systems mentioned above in the text; see also fn. 14. See also Jonas Lerman,
‘Big Data and Its Exclusions’ (2013) 66 Stan. L. Rev. Online 55.
87
See Robinson + Yu (n 14) 6.
88
This is the case of the errors that affect the E-Verify system,
which is used in the US to verify if a new worker is legally eligible to work in the US. See Robinson + Yu (n 14) 12–14; National
Immigration Law Center, ‘Verification Nation’ (2013) 6 <www.nilc.org/
document.html?id=957> accessed 29 January 2015.
89
See also Gandy (n 16) 1099–1100.
90
See above para 1.
91
See Jeremy Gorner, ‘Chicago police use ‘heat list’ as strategy to
prevent violence. Officials generate analysis to predict who will likely
be involved in crime, as perpetrator or victim, and go door to door
to issue warnings’ Chicago Tribune (Chicago, 21 August 2013) <http://
articles.chicagotribune.com/2013-08-21/news/ct-met-heat-list
-20130821_1_chicago-police-commander-andrew-papachristos
-heat-list> accessed 25 February 2015.
92
See U.S. Department of Justice – Criminal Division, Office of the
Assistant Attorney General, ‘Annual letter’ (2014) 6–7, 13 <http://
www.justice.gov/criminal/foia/docs/2014annual-letter-final
-072814.pdf> accessed 29 January 2015 (“This phenomenon ultimately raises constitutional questions because of the use of
groupbased characteristics and suspect classifications in the analytics. Criminal accountability should be primarily about prior bad
acts proven by the government before a court of law and not some
future bad behavior predicted to occur by a risk assessment instrument. Second, experience and analysis of current risk
assessment tools demonstrate that utilizing such tools for
determining prison sentences to be served will have a disparate
and adverse impact on offenders from poor communities already
struggling with many social ills”). See also Administrative Office
of the United States Courts – Office of Probation and Pretrial Services, ‘An Overview of the Federal Post Conviction Risk Assessment’
(2011) <http://www.uscourts.gov/uscourts/FederalCourts/PPS/
PCRA_Sep_2011.pdf> accessed 29 January 2015; Underwood (n 7)
1409–1413.
248
Discrimination – the different treatment of different situations – also appears in commercial contexts to offer tailored
services to consumers. In this case, where the interests are of
a purely private nature, commercial practices may lead to price
discrimination93 or the adoption of differential terms and conditions depending on the assignment of consumers to a specific
cluster.94
Thus consumers classified as “financially challenged” belong
to a cluster “[i]n the prime working years of their lives [. . .] including many single parents, struggl[ing] with some of the
lowest incomes and little accumulation of wealth”. This implies
the following predictive viewpoint, based on big data analytics and regarding all consumers in the cluster: “[n]ot particularly
loyal to any one financial institution, [and] they feel uncomfortable borrowing money and believe they are better off having
what they want today as they never know what tomorrow will
bring”.95 It is not hard to imagine the potential discriminatory consequences of similar classifications with regard to
individuals and groups.
It should be noted that these forms of discrimination are
not necessarily against the law, especially when they are not
based on individual profiles and only indirectly affect individuals
93
Price discrimination or “differential pricing” is the practice of
charging customers different prices for the same product; see Executive Office of the President of the United States-Council of
Economic Advisers (n 68), 4–5. The cases considered in this article
mainly concern the so-called third-degree price differentiation,
which occurs when sellers charge different prices to different segments of the market. See also Alex Rosenblat, Rob Randhava, danah
boyd, Seeta Peña Gangadharan, and Corrine Yu Produced, ‘Data &
Civil Rights: Consumer Finance Primer’ (2014) <http://www
.datacivilrights.org/pubs/2014-1030/Finance.pdf> accessed 15 March
2015.
94
See Executive Office of the President of the United StatesCouncil of Economic Advisers (n 68); Federal Trade Commission (n
2) 3, 19–21; Dixon and Gellman (n 9). See also Timothy C. Lambert,
‘Fair Marketing: Challenging Pre-Application Lending Practices’ (1999)
87 Geo. L. J. 2182.
95
See Federal Trade Commission (n 2) 20, fn. 52.
as part of a category, without their direct identification.96 For
this reason, existing legal provisions against individual discrimination might not be effective in preventing the negative
outcomes of these practices, if adopted on a collective basis.
Still, such cases clearly show the importance of the collective
dimension of the use of information about groups of individuals.
Within the EU, such data analysis focussing on clustered
individuals may not represent a form of personal data
processing,97 since the categorical analytics methodology does
not necessarily make it possible to identify a person.98 Moreover,
96
Article 4 (12b) of the EU Proposal defines “profile” as “set of data
characterising a category of individuals that is intended to be applied
to a natural person”. See also Articles 4 (12a), 14 (1a) (h) and 20,
EU Proposal; Article 29 Data Protection Working Party, ‘Advice paper
on essential elements of a definition and a provision on profiling
within the EU General Data Protection Regulation’ <http://
ec.europa.eu/justice/data-protection/article-29/documentation/
other-document/files/2013/20130513_advice-paper-on
-profiling_en.pdf> accessed 29 March 2015; Article 29 Data Protection Working Party, ‘Opinion 01/2012 on the data protection reform
proposals’ (2012) <http://ec.europa.eu/justice/data-protection/
article-29/documentation/opinion-recommendation/files/2012/
wp191_en.pdf> accessed 29 March 2015. Regarding the decisions
that affect an individual as member of a specific cluster of people,
it should be noted that in many cases these decisions are not based
solely on automated processing; see Zarsky (n 17) 1518–1519. In this
sense, credit scoring systems have reduced but not removed human
intervention on credit evaluation. At the same time, classifications often regard identified or identifiable individuals. See Article
29 Data Protection Working Party, ‘Opinion 2/2010 on online
behavioural advertising’ (2010) <http://ec.europa.eu/justice/policies/
privacy/docs/wpdocs/2010/wp171_en.pdf> accessed 29 March 2015;
Data Protection Working Party, ‘Working Document 02/2013 providing guidance on obtaining consent for cookies’ (2013) 5–6 <http://
ec.europa.eu/justice/data-protection/article-29/documentation/
opinion-recommendation/files/2013/wp208_en.pdf> accessed 29
March 2015. Regarding the applicability of the Data Protection Directive in case of automated profiling, see Lee A. Bygrave, ‘Minding
the Machine: Article 15 of the EC Data Protection Directive and Automated Profiling’ in this Review (2001), vol. 17, issue 1, 17–24;
Schreurs, Hildebrandt, Kindt and Vanfleteren (n 82) 241–257; Judith
Rauhofer, ‘Round and Round the Garden? Big Data, Small Government and the Balance of Power in the Information Age’ (2014)
University of Edinburgh School of Law, Research Paper Series 2014/
06, 5, 10 <http://papers.ssrn.com/sol3/papers.cfm?abstract
_id=2389981> accessed 15 March 2015. See also Valeria Ferraris et
al., ‘Working Paper. Defining Profiling’ (2014) 15–20 <http://
www.unicri.it/special_topics/citizen_profiling/WP1_final
_version_9_gennaio.pdf> accessed 15 June 2015; Douwe Korff, ‘Data
Protection Laws in the EU: The Difficulties in Meeting the Challenges Posed by Global Social and Technical Developments’ (2010)
82–89 <http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1638949>
accessed 7 April 2015.
97
See Article 29 Data Protection Working Party, ‘Opinion 4/2007
on the concept of personal data’ (2007) <http://ec.europa.eu/justice/
policies/privacy/docs/wpdocs/2007/wp136_en.pdf> accessed 25
January 2015.
98
See also above fn. 96. On the blurring of the border between
group profiles and personalised profiles, see also Hildebrandt, ‘Profiling: From Data to Knowledge. The challenges of a crucial
technology’ (n 6).
group profiles can be made using anonymised data.99 This
reduces the chances of individuals taking action against biased
representations of themselves within a group or having access
to the data processing mechanisms, since the anonymised information used for group profiling cannot be linked to them.100
Even so, group profiling does make it possible to take decisions affecting a multiplicity of individuals.101 In this sense, the
main target of the collective dimension of data processing is
not the data subject, but the clusters of people created by big
data gatherers.
The interests that assume relevance therefore have a supraindividual nature and a collective dimension,102 which are not
adequately addressed by the existing data protection legal
framework. These interests may be shared by an entire group
without conflicts between the views of its members (aggregative
interests) or with conflicts between the opinions of its members
99
On the limits of anonymisation in the big data context, see
Arvind Narayanan, Joanna Huey, Edward W. Felten, ‘A Precautionary Approach to Big Data Privacy’ (2015) <http://randomwalker.info/
publications/precautionary.pdf> accessed 4 April 2015; Arvind
Narayanan, Edward W. Felten, ‘No silver bullet: De-identification
still doesn’t work’ (2014) <http://randomwalker.info/publications/
no-silver-bullet-de-identification.pdf> accessed 25 March 2015; Ohm
(n 67); United States General Accounting Office, ‘Record Linkage and
Privacy. Issues in creating New Federal Research and Statistical Information’ (2011) 68–72 <http://www.gao.gov/assets/210/201699.pdf>
accessed 14 December 2013; Golle (n 67); Latanya Sweeney, ‘Only
You, Your Doctor, and Many Others May Know’ Technology Science.
2015092903. September 29, 2015 <http://techscience.org/a/
2015092903> accessed 28 November 2015; Sweeney, ‘Simple
Demographics Often Identify People Uniquely’ (n 67); Sweeney,
‘Foundations of Privacy Protection from a Computer Science Perspective’ (n 67).
100
and Limits (n 34) 319; Schreurs, Hildebrandt, Kindt and Vanfleteren
(n 82) 252–253; Rauhofer (n 96). With regard to the EU Proposal, see
also Bert-Jaap Koops, ‘The trouble with European data protection
law’ (2014) 4(4) Int’l. Data Privacy Law 257–258. But see above
fn. 99.
101
This happens, for instance, in the management of smart cities
or in the decisions adopted on the basis of credit scoring systems.
Against this background, Mireille Hildebrandt observed that “once
a profile is linked to an identifiable person – for instance in the
case of credit scoring – it may turn into a personal data, thus reviving the applicability of data protection legislation”, see
Hildebrandt, ‘Profiling: From Data to Knowledge. The challenges of
a crucial technology’ (n 6) 550. See also Vedder (n 9) (“Categorial
privacy can be considered as relating to information (1) which was
originally taken from the personal sphere of individuals, but which,
after aggregation and processing according to statistical methods,
is no longer accompanied by identifiers indicating individual natural
persons, but, instead, by identifiers of groups of persons, and
(2) which, when attached to identifiers of groups and when disclosed, is apt to cause the same kind of negative consequences to
the members of those groups as it would for an individual person
if the information were accompanied by identifiers of that
individual”).
102
See Newman (n 76) 131.
249
(non-aggregative interests).103 If the group is characterised by
non-aggregative interests, the collective nature of the interest is represented by the fundamental values of a given society
(e.g. environmental protection).
The notion of collective non-aggregative interests seems to
be the best way to describe the collective dimension of data
protection, which becomes important in the discrimination
cases mentioned above. Although individuals may have different opinions about the balance between the conflicting
interests,104 there are some collective priorities concerning
privacy and data-protection that are of relevance to the general
interest. Here the rationale for collective data protection is
mainly focussed on the potential harm to groups caused by
extensive and invasive data processing.
3.2.
Collective interests in data protection and their
representation
Privacy and data protection are context-dependent notions,
which vary from culture to culture and across historical
periods.105 In the same way, the related collective dimensions
are necessarily influenced by historical and geographical variables and are the result of action by policymakers. For these
reasons, it is impossible to define a common and fixed balance
between collective data protection and conflicting interests.
There are jurisdictions that give greater priority to national and security interests, which in many cases prevail over
individual and collective data protection; meanwhile, in some
countries extensive forms of social surveillance are considered disproportionate and invasive. Therefore, any balancing
test must focus on a specific social context in a given historical moment.106 As has been pointed out in the literature,107
defining prescriptive ethical guidelines concerning the values
that should govern the use of big data analytics and the related
balance of interests is problematic.
Given such variability, from a theoretical perspective a
common framework for a balancing test can be found in the
values recognised by international charters of fundamental
103
Newman (n 76) 131–132 makes this distinction and defines these
two categories of interests respectively as “shared” and “collective” interests. As observed by Finnis, a collective interest in which
the conflict is diminished may become a shared interest. See John
Finnis, ‘The Authority of Law in the Predicament of Contemporary Social Theory’ (1984) 1 J.L. Ethics & Pub. Pol’y 115, 135–136.
104
In this sense, an extensive group profiling for commercial purposes can be passively accepted, considered with favour or perceived
as invasive and potentially discriminatory. The same divergence
of opinions and interests exists with regard to government social
surveillance for crime prevention and national security, where part
of the population is in favour of surveillance, due to concerns about
crime and terrorism.
105
See Westin (n 25) 183–187; Whitman (n 48); Bygrave, Data Protection Law. Approaching Its Rationale, Logic and Limits (n 34) 327;
Nissenbaum (n 72); Irwin Altman, ‘Privacy Regulation: Culturally
Universal or Culturally Specific?’ (1977) Journal of Social Issues 33(3)
66–84.
106
See in this sense the different attitudes of U.S. government with
regard to surveillance, before and after the September 11, 2001, terrorist attacks. See also Bygrave, ‘Privacy Protection in a Global
Context. A Comparative Overview’ (n 48) 329.
107
See Wright (n 72) 200.
250
rights. These charters provide a baseline from which to identify the values that can serve to provide ethical guidance
and define the existing relationships between these
values.108
In addition, the context-dependent framework of values and
the relationship between conflicting interests and rights needs
to be specified with regard to the actual use of big data analytics. In Europe, for instance, commercial interests related to
credit score systems can generally be considered compatible
with the processing of personal information, providing that the
data are adequate, relevant and not excessive in relation to the
purposes for which it is collected.109 Even so, specific big data
analytics solutions adopted by some companies for credit
scoring purposes may lead to a disproportionate scrutiny of
a consumer’s private life. The same reasoning can also be
applied to smart mobility solutions, which can potentially lead
to extensive social surveillance. This means a prior case-bycase risk-assessment is necessary to mitigate the potential
impact of these solutions on data protection and individual
freedoms.110
This “in-context” balance of conflicting interests based on
an impact assessment of complex data collection and processing systems,111 should not be conducted by consumers or
companies, but must entail the active involvement of various
stakeholders. Against this background, an important aspect of
the protection of collective interests relating to personal information is an analysis of the existing conflicting interests
and the representation of the issues regarding the individuals grouped in clusters by the data gatherers.112
Here it is useful to briefly consider the fields in which the
group dimension of data protection is already known in more
traditional contexts that are not characterised by extensive data
collection and use of analytics. For instance, labour law
recognises this collective dimension of rights and the dualism
between individuals and groups.113 Under certain circumstances, trade unions and employees’ representatives may
concur in taking decisions that affect the employees and have
an impact on data protection in the workplace.114
Collective agreement on these decisions is based on the recognition that the power imbalance in the workplace means that,
in some cases, the employee is unaware of the implications
of employer’s policies (e.g. employers’ workplace surveillance practices). Moreover, in many cases, this imbalance makes
it difficult for employees to object to the illegitimate processing of their data.
Entities representing collective interests (e.g. trade unions)
are less vulnerable to power imbalance and have a broader
vision of the impact of the employer’s policies and decisions.
It should also be noted that the employer’s unfair policies and
forms of control are often oriented towards discriminatory measures that affect individual workers, even though they are
targeted at the whole group.
This collective representation of common interests is also
adopted in other fields, such as consumer protection and environmental protection. These contexts are all characterised
by a power imbalance affecting one of the parties directly involved (employees, consumers or citizens). Furthermore, in
many cases the conflicting interests refer to contexts where
the use of new technologies makes it hard for users to be aware
of the potential negative implications.
The same situation of imbalance often exists in the big data
context, where data subjects are not in a position to object to
the discriminatory use of personal information by data
108
See Wright (n 72) 201–202.
See Articles 18 and 20 of the Directive 2014/17/EU. See also Article
8 of the Directive 2008/48/EC on credit agreements for consumers
and repealing Council Directive 87/102/EEC.
110
See e.g. Italian Data Protection Authority (Autorità Garante per
la protezione dei dati perosnali, hereafter Gar.), 6 September 2006
(decision, doc. web n. 1339531) <http://www.garanteprivacy.it/web/
guest/home/docweb/-/docweb-display/docweb/1339531> (text in
Italian); Gar., 28 December 2006 (decision, doc. web n. 1413380)
<http://www.garanteprivacy.it/web/guest/home/docweb/-/docweb
-display/docweb/1413380> (text in Italian); Gar., 6 September 2006
(decision, doc. web n. 1339692) <http://www.garanteprivacy.it/web/
guest/home/docweb/-/docweb-display/docweb/1339692> (text in
Italian). See also Alessandro Mantelero, ‘Data protection, e-ticketing
and intelligent systems for public transport’ (2015) 5(4) Int’l. Data
Privacy Law 309–320.
111
Moreover, these systems are influenced by lock-in effects. There
are two different kinds of lock-ins: technological lock-in and social
lock-in. The first is related to the technological standards and data
formats that are adopted by service providers. This lock-in represents a limit to data portability and migration from one service to
another. The second lock-in (social lock-in) is related to the dominant position held by some big players. This lock-in is evident, for
example, in the social networks market, where there is an incentive to remain on a network, given the numbers of social
relationships created by users.
109
112
On group right, see also Julia Stapleton, Group rights: perspectives since 1900 (Thoemmes Press 1995); Will Kymlicka, ‘Individual
and Community Rights’ in Judith Baker (ed), Group Rights (University of Toronto Press 1994) 17–33; Eugene Schlossberger, A Holistic
Approach to Rights: Affirmative Action, Reproductive Rights, Censorship, and Future Generations (University Press of America 2008)
ch. 10; Helen O’Nions, Minority Rights Protection in International Law:
The Roma of Europe (Ashgate Publishing 2007) 26–28; Peter Jones,
‘Group Rights’ (2008) in Stanford Encyclopedia of Philosophy http://
stanford.library.usyd.edu.au/entries/rights-group/ accessed 4 July
2015; Newman (n 76); Will Kymlicka, ‘The New Debate Over Minority Rights’ in Wayne Norman and Ronald Beiner (eds.) Canadian
Political Philosophy: Contemporary Reflections (Oxford University Press
2000) 159–176; John Packer, ‘Problems in Defining Minorities’ in
Deirdre Fottrell and Bill Bowring (eds), Minority and group rights in
the new millennium (M. Nijhoff Publishers 1999) 241 ff; Will Kymlicka,
‘Individual and Community Rights’ in Judith Baker (ed) Group Rights
(University of Toronto Press 1994) 17–33.
113
See Italian Articles 4 and 8, Act 300, 20 May 1970 (Statute of the
Workers’ Rights).
114
See above at fn. 52. See also Bygrave and Wiese Schartum (n
41) 170.
gatherers.115 Data subjects often do not know the basic steps
of data processing,116 and the complexity of the process means
that they are unable to negotiate their information and are not
aware of the potential collective prejudices that underlay its
use.117 This is why it is important to recognise the role of entities representing collective interests, as happens in the above
cases.
Employees are part of a specific group, defined by their relationship with a single employer; therefore, they are aware
of their common identity and have mutual relationships. By
contrast, in the big data context, the common attributes of the
group often only become evident in the hands of the data
gatherer.118
Data subjects are not aware of the identity of the other
members of the group, have no relationship with them and
have a limited perception of their collective issues. Furthermore, these groups shaped by analytics have a variable
geometry and individuals can shift from one group to another.
This does not undermine the idea of a representing collective data protection interests. On the contrary, this atomistic
dimension makes the need for collective representation more
urgent. However, it is hard to imagine representatives appointed by the members of these groups, as is instead the case
in the workplace.
In this sense there are similarities with consumer law, where
there are collective interests (e.g. product security, fair commercial practices), but the potential victims of harm have no
relationship to one another. Thus, individual legal remedies
115
In the digital economy, consumers often accept not having an
effective negotiation of their personal information, due to market
concentration and related social and technological lock-ins. See
above fn. 111.
116
See also Alessandro Acquisti, Laura Brandimarte and George
Loewenstein, ‘Privacy and human behavior in the age of information’ (2015) 347(6221) Science 509–514.
117
The complexity of data processing in the big data environment does not offer users a real chance to understand it and make
their choice. See Pasquale (n 9) 143–144; Laura Brandimarte,
Alessandro Acquisti, and George Loewenstein, ‘Misplaced Confidences: Privacy and the Control Paradox’ (2010), Ninth Annual
Workshop on the Economics of Information Security <http://
www.heinz.cmu.edu/~acquisti/papers/acquisti-SPPS.pdf> accessed
27 February 2014; Joseph Turow, Chris Jay Hoofnagle, Deirdre K. Mulligan, and Nathaniel Good, ‘The Federal Trade Commission and
Consumer Privacy in the Coming Decade’ (2007) 3 ISJLP 723–749
<http://scholarship.law.berkeley.edu/facpubs/935> accessed 27 February 2014; Federal Trade Commission (n 2) 42. On the limits of the
traditional notices, see also Rayan M. Calo, ‘Against Notice Skepticism in Privacy (and Elsewhere)’ (2013) 87(3) Notre Dame L. Rev.
1027, 1050–1055 <http://scholarship.law.nd.edu/ndlr/vol87/iss3/3>
accessed 27 February 2014; Daniel J. Solove, ‘Introduction: Privacy
Self-management and The Consent Dilemma’ (2013) 126 Harv. L.
Rev. 1880, 1883–1888; World Economic Forum, ‘Unlocking the Value
of Personal Data: From Collection to Usage’ (2013) 18 <http://
www3.weforum.org/docs/WEF_IT_UnlockingValuePersonalData
_CollectionUsage_Report_2013.pdf> accessed 27 February 2014.
118
See also Bygrave, Data Protection Law. Approaching Its Rationale,
Logic and Limits (n 34) 283–284.
251
must be combined with collective remedies.119 Examples of possible complementary solutions are provided by consumer law,
where independent authorities responsible for consumer protection, class action lawsuits and consumer associations play
an important role.
In the field of big data analytics, the partially hidden nature
of the processes and their complexity probably make timely
class actions more difficult than in other fields. For instance,
in the case of a product liability, the damages are often more
evident making it easier for the injured people to react.120 On
the other hand, associations that protect collective interests
can play an active role in facilitating reaction to unfair practices and, moreover, they can be involved in a multi-stakeholder
risk assessment of the specific use of big data analytics.121
The involvement of such bodies requires specific procedural criteria to define the entities that may act in the collective
interest.122 This is more difficult in the context of big data, where
the groups created by data gatherers do not have a stable character. In this case, an assessment of the social and ethical
impact of analytics often provides the opportunity to discover how data processing affects collective interests and thus
identify the potential stakeholders.123
3.3.
The role of data protection authorities
How collective interests should be protected against discrimination and social surveillance in the use of big data analytics
is largely a matter for the policymakers. Different legal systems
and different balances between the components of society
suggest differing solutions. Identifying the independent authority charged with protecting collective interests may
therefore be difficult.
119
The same approach has been adopted in the realm of antidiscrimination laws; see European Commission, ‘Developing AntiDiscrimination Law in Europe. The 28 EU Member States, the Former
Yugoslav Republic of Macedonia, Iceland, Liechtenstein, Norway and
Turkey compared’ (n 82) 83–110. See also Lilla Farkas, ‘Collective
actions under European anti-discrimination law’ (2014) 19 European Anti-Discrimination Law Rew. 25–40.
120
As demonstrated by recent revelations on NSA case, in the big
data context people are not usually aware of being under surveillance. Only leaks of information can disclose these practices, open
a debate on their legitimacy and give the chance to individuals to
bring legal actions. See also European Parliament, ‘Resolution of
4 July 2013 on the US National Security Agency surveillance programme, surveillance bodies in various Member States and their
impact on EU citizens’ privacy’ (2013) <http://www.europarl
.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+TA+P7-TA-20130322+0+DOC+XML+V0//EN> accessed 27 February 2014. On the role
played by group actions in order to protect individual and collective interests concerning personal information, see Bygrave, Data
Protection Law. Approaching Its Rationale, Logic and Limits (n 34), 288–290.
121
See also para 3.3.
122
See art. 76 of the EU Proposal.
123
For these reasons, a preventive approach based on risk assessment seems to be more effective than ex post legal actions. Moreover,
it also contributes to tackle the risks of hidden forms of data processing, which often create an asymmetric distribution of the control
over information in our society. See also Alessandro Mantelero,
‘Social Control, Transparency, and Participation in the Big Data World’
(2014) April Journal of Internet Law 23–29.
252
Many countries have independent bodies responsible for supervising specific social surveillance activities, and other bodies
focused on anti-discrimination actions.124 In other countries,
this responsibility is spread across various authorities, which
take different approaches, use different remedies and do not
necessarily cooperate in solving cases with multiple impacts.
Meanwhile, a central element in the risk-assessment of big
data analytics is the analysis of data processing, the factor
common to all these situations, regardless of the potential harm
to collective interests. For this reason, data protection authorities can play a key role in the risk assessment processes, even
if they are not focused on the specific social implications (e.g.
discrimination).
On the other hand, if we take a different approach that takes
into consideration the various negative effects generated by
the use of big data (discrimination, unfair consumer practices, social control, etc.), we should involve multiple entities
and authorities. Nevertheless, as we have seen the end result
may be a fragmented and potentially conflicting decisionmaking process that may underestimate the use of data, which
is the common core of all these situations.125
Furthermore, the data protection authorities are accustomed to addressing collective issues and have already
demonstrated that they do consider both the individual and
the wider collective dimension of data processing.126 Focusing on data protection and fundamental rights, they are also
well placed to balance the conflicting interests around the use
of data.
The adequacy of the solution is also empirically demonstrated by important cases decided by data protection
authorities concerning data processing projects with significant
124
See European Commission, ‘Developing Anti-Discrimination Law
in Europe. The 28 EU Member States, the Former Yugoslav Republic of Macedonia, Iceland, Liechtenstein, Norway and Turkey
compared’ (n 82) 113–125.
125
See also Lerman (n 86) 60, who points out the limits of the U.S.
equal protection doctrine in the context of big data analytics.
126
See fn. 52. See also e.g. Article 29 Data Protection Working Party,
‘Explanatory Document on the Processor Binding Corporate Rules’
(2015) <http://ec.europa.eu/justice/data-protection/article-29/
documentation/opinion-recommendation/files/2015/wp204
.rev_en.pdf> accessed 29 November 2015; Article 29 Data Protection Working Party, ‘Opinion 9/2014 on the application of Directive
2002/58/EC to device fingerprinting’ (2014) <http://ec.europa.eu/
justice/data-protection/article-29/documentation/opinion
-recommendation/files/2014/wp224_en.pdf>; Article 29 Data Protection Working Party, ‘Opinion 8/2014 on the Recent Developments
on the Internet of Things’ (2014) <http://ec.europa.eu/justice/data
-protection/article-29/documentation/opinion-recommendation/
files/2014/wp223_en.pdf>; Article 29 Data Protection Working Party,
‘Opinion 03/2012 on developments in biometric technologies’ (2012)
<http://ec.europa.eu/justice/data-protection/article-29/
documentation/opinion-recommendation/files/2012/wp193
_en.pdf>; Article 29 Data Protection Working Party, ‘Opinion 02/
2012 on facial recognition in online and mobile services’ (2012)
<http://ec.europa.eu/justice/data-protection/article-29/
documentation/opinion-recommendation/files/2012/
wp192_en.pdf>.
social and ethical impacts.127 These cases show that decisions to assess the impact of innovative products, services and
business solutions on data protection and society are not normally on the initiative of the data subjects, but primarily on
that of the data protection authorities who are aware of the
potential risks of such innovations. Based on their balancing
tests, these authorities are in a position to suggest measures
that companies should adopt to reduce the risks discussed here
and to place these aspects within the more general framework of the rights of the individual, as a single person and as
a member of a democratic society.
The risk assessment represents the opportunity for group
issues to be identified and addressed. Thus, bodies representing collective interests should not only partially exercise
traditional individual rights on behalf of data subjects,128 but
also exercise other autonomous rights relating to the collective dimension of data protection. These new rights mainly
concern participation in the risk assessment process, which
should take a multi-stakeholder approach.129
Against this background, data protection authorities may
involve in the assessment process the various stakeholders,
which represent the collective interests affected by specific data
processing projects.130 This would lead to the definition of a
new model in which companies that intend to use big data analytics would undergo an assessment prior to collecting and
processing data.
127
See above fns. 52 and 126.
The stakeholders may have right of access to the documents
that describe the specific structure and general purposes of big data
processing. However, in order to protect the legitimate interests of
companies and governments, the data protection authorities might
limit this disclosure to third parties. See also art. 76 of the EU Proposal and Bygrave, Data Protection Law. Approaching Its Rationale, Logic
and Limits (n 34) 274–282.
129
Note that the extent of the rights conferred upon the different stakeholders in the protection of collective privacy is largely
a matter for policymakers to decide and would depend on the nature
and values of the different socio-legal contexts.
130
See also Wright (n 72) 201–202, 215–220; Danielle Keats Citron,
‘Technological Due Process’ (2008) 85(6) Wash. U. L. Rev. 1249, 1312.
A different assessment exclusively based on the adoption of security standards or corporate self-regulation would not have the
same extent and independency. This does not mean that, in this
framework, forms of standardisation or co-regulation cannot be
adopted.
128
253
The assessment would not only focus on data security and
data protection,131 but also consider the social and ethical
impacts relating to the collective dimension of data use in a
given project.132 This assessment should be conducted by third
parties and supervised by the data protection authorities.133
Once this multiple-impact assessment is approved by the data
protection authorities, the ensuing data processing would be
considered secure in protecting personal information and collective interests.134
Although data protection authorities are already engaged to
some degree in addressing the collective dimension,135 the suggested solution would lead to a broader and deeper assessment,
which would become mandatory.136 This proposal is therefore
in line with the view that a licensing scheme might “prove to
be the most effective means of ensuring that data protection
principles do not remain ‘law-in-book’ with respect to profiling practices”.137 Finally, it should be noted that a different risk
assessment model, which also takes into account the ethical
and social effects of data use, directly affects data processing
131
133
On traditional forms of privacy impact assessment, see Roger
Clarke, ‘Privacy impact assessment: Its origins and development’
in this Review 2009, vol. 25, issue 2, 123–129; David Flaherty, ‘Privacy
impact assessments: an essential tool for data protection’ (2000)
7(5) Priv. Law & Pol’y Rep. 45; David Wright, ‘The state of the art
in privacy impact assessment’ in this Review 2012, vol. 28, issue 1
54–61; David Wright and Paul De Hert (eds), Privacy Impact Assessment (Springer 2012); David Wright, Michael Friedewald, and Raphael
Gellert, ‘Developing and Testing a Surveillance Impact Assessment Methodology’ (2015) 5(1) Int’l. Data Privacy Law 40–53. See
also Commission nationale de l’informatique et des libertés, ‘Étude
d’impact sur la vie privée (EIVP)’ (2015) <http://www.cnil.fr/fileadmin/
documents/Guides_pratiques/CNIL-PIA-1-Methode.pdf> accessed
16 July 2015.
132
In the big data context, another important aspect is the transparency of the algorithms used by companies. See Citron and
Pasquale (n 9) 5, 10–11, 25, 31; Pasquale (n 9) 193, 216–218. See also
Viktor Mayer-Schönberger and Kenneth Cukier, Big Data. A Revolution That Will Transform How We Live, Work and Think (n 12) 179–
182; they suggest a model based on independent internal and
external audits. A wider access to the logic of the algorithms was
required by Article 29 Data Protection Working Party, ‘Opinion 03/
2013 on purpose limitation’ (2013) 47 <http://ec.europa.eu/justice/
data-protection/article-29/documentation/opinion-recommendation/
files/2013/wp203_en.pdf> accessed 27 February 2014. See also
Tarleton Gillespie, ‘The Relevance of Algorithms’ in Tarleton Gillespie, Pablo J. Boczkowski, and Kirsten A. Foot (eds) Media
Technologies. Essays on Communication, Materiality, and Society (MIT
Press 2014) 167, 194; Dixon and Gellman (n 9), 7 (“Trade secrets have
a place, but secrecy that hides racism, denies due process, undermines privacy rights, or prevents justice does not belong anywhere”).
But see Recital n. 51 of EU Proposal, text adopted by the Council
of the European Union, Brussels, 19 December 2014 (“Every data
subject should therefore have the right to know and obtain communication in particular for what purposes the data are processed
[. . .], what is the logic involved in any automatic data processing
and what might be, at least when based on profiling, the consequences of such processing. This right should not adversely affect
the rights and freedoms of others, including trade secrets or intellectual property and in particular the copyright protecting the
software”). On the interest in knowing the logic of profiling, see
also Schreurs, Hildebrandt, Kindt and Vanfleteren (n 82) 253–256.
On transparency in decisional processes based on big data analytics, see also Zarsky (n 17) 1523–1530.
The entire system will work only if the political and financial
autonomy of data protection authorities from governments and corporations is guaranteed. Moreover, data protection authorities would
need new competence and resources in order to bear the burden
of the supervision and approval of these multiple-impact assessments. For these reasons, a model based on mandatory fees – paid
by companies when they submit their requests for authorisation
to data protection authorities – would be preferable. See Mantelero,
‘The future of consumer data protection in the E.U. Rethinking the
“notice and consent” paradigm in the new era of predictive analytics’ (n 20). It should also be noted that, in cases of large scale
and multinational data collection, forms of mutual assistance and
cooperation may facilitate the role played by data protection authorities in addressing the problems related to the dimensions of
data collection and data gatherers. See also Gwendal Le Grand and
Emilie Barrau, ‘Prior Checking, a Forerunner to Privacy Impact Assessments’ in Wright and De Hert (n 131) 112–116.
134
Therefore, in this scenario, companies can enlist users in the
data processing without any prior consent, provided they give notice
of the results of the assessment and provide an opt-out option.
See more extensively Mantelero, ‘The future of consumer data protection in the E.U. Rethinking the “notice and consent” paradigm
in the new era of predictive analytics’ (n 20), 654–659. In this case,
although this assessment represents an economic burden for companies, it allows those who pass to use data for complex and
multiple purposes, without requiring users to opt-in. At the same
time, from the users’ side, the assessment supervised by data protection authorities provides an effective evaluation of risks, while
the option to opt-out allows users to choose to not be a part of the
data collection. See also Citron and Pasquale (n 9) 24–28. The suggested model represents a significant change in the traditional
approach to data protection, but this is in line with the approach
adopted in other fields characterised by the presence of risks for
individuals and society (e.g. authorisation procedure for human
medicines, mandatory security standards provided by product liability laws, security standards for industrial activities). For this
reason, it would be necessary to adopt a subset of rules for big data
analytics, which focuses on multiple risk assessment and a deeper
level of control by data protection authorities.
135
See above fns. 52 and 126.
136
See fns. 133, 134.
137
and Limits (n 34), 373–374. See also Lee A. Bygrave, Data Privacy Law.
An International Perspective (Oxford University Press 2014) 183–186.
254
design. Literature on privacy by design138 has clearly stressed
the relevance of embedding privacy values in the device and services architecture. To achieve this goal, a preliminary analysis
of the use of personal information in each specific application
(good or service) is required to shape this use according to data
protection purposes. Based on this assessment, suitable measures would be taken to reduce the potential negative outcomes
of data use.
This strict relationship between risk assessment and solutions by design implies that any change in the nature of the
assessment would affect the architectural solutions adopted.
Thus, the multiple impact assessment suggested would necessarily lead companies to consider a broader range of bydesign solutions to mitigate the additional ethical and social
concerns.139
4.
Conclusions
This article does not provide complete answers to the various
issues concerning the collective dimension of privacy and data
protection relating to the use of big data analytics. The analysis offers an introductory study of a new approach to group
privacy that would appear necessary to adequately consider
the non-aggregative interests arising from the data-driven
society.
138
See Ann Cavoukian, ‘Privacy by design. From rhetoric to reality’
(2014) 12–18, 65–100 <http://www.ipc.on.ca/images/Resources/
PbDBook-From-Rhetoric-to-Reality.pdf> accessed 27 February 2014;
Ann Cavoukian, ‘Privacy by Design: Leadership, Methods, and
Results’ in Serge Gutwirth, Ronald Leenes, Paul De Hert, Yves Poullet
(eds), European Data Protection: Coming of Age (Springer 2013) 175–
202; Ann Cavoukian, ‘Privacy by Design: Origins, Meaning, and
Prospects for Assuring Privacy and Trust in the Information Era’
in Yee, G O M (ed), Privacy Protection Measures and Technologies in Business Organizations: Aspects and Standards (IGI Global 2012) 170–208;
Ira S. Rubenstein, ‘Regulating Privacy By Design’ (2011) 26 Berkeley Tech. L. J. 1409–1456; Peter Schaar, ‘Privacy by Design’ (2010) 3(2)
Identity in the Information Society 267–274. See also Article 29 Data
Protection Working Party, ‘Opinion 07/2013 on the Data Protection Impact Assessment Template for Smart Grid and Smart
Metering Systems (‘DPIA Template’) prepared by Expert Group 2 of
the Commission’s Smart Grid Task Force’ (2013) <http://ec.europa.eu/
justice/dataprotection/article-29/documentation/opinion
-recommendation/files/2013/wp209_en.pdf> accessed 27 February 2014; Article 29 Data Protection Working Party, ‘Opinion 9/2011
on the revised Industry Proposal for a Privacy and Data Protection Impact Assessment Framework for RFID Applications’ (2011)
<http://ec.europa.eu/justice/policies/privacy/docs/wpdocs/2011/
wp180_en.pdf> accessed 27 February 2014; Woodrow Hartzog and
Frederic Stutzman, ‘Obscurity by Design’ (2013) 88 Wash. L. Rev. 385,
397; Federal Trade Commission, ‘Protecting Consumer Privacy in
an Era of Rapid Change. Recommendations for Business and
Policymakers’ (2012) 22–24 <http://www.ftc.gov/sites/default/
files/documents/reports/federal-trade-commission-report
-protectingconsumer-privacy-era-rapid-change-recommendations/
120326privacyreport.pdf> accessed 25 June 2014.
139
See Wright (n 72).
Big data analytics are creating a new digital landscape that
cannot be described as a mere increase in the quantity of information processed. The predictive nature of the inferences
extracted from databases, the complexity of data processing
and its obscurity, as well as the categorical approach, distinguish it from previous profiling solutions.
From the group perspective, big data analytics create new
kind of groups, which cannot be compared with the traditional concept of a group. They are the result of aggregations
of information produced by data gatherers and have a variable geometry, setting them apart from the previous static
categories used for group profiling. Moreover, data subjects are
not aware of the identity of the other members of the group
and have a limited or no perception of their collective issues,
whereas in traditional groups there is an awareness of being
part of a group and groups have external visibility.
The new scale of data processing, the pervasive diffusion
of data-based applications, the evolution and complexity of
group profiling represent important changes with respect to
the previous scenario.140 At the dawn of the data-driven society,
the question arises whether it is necessary to reconsider the
traditional approach to group privacy and data protection, which
is mainly based on the model of individual rights.
This article gives an affirmative response to this question
on the basis of the impact that big data analytics have on data
processing and data-driven decisions. The shift in the data processing paradigm and the new forms of categorical approach
have a disruptive effect on the traditional idea of group privacy
and highlight its limits.
The new scale entails the recognition of a new layer, represented by groups’ need for the protection of their collective
data protection rights. In this scenario, data protection concerns not only individuals, but also the collective dimension,
associated with potential harm to groups in terms of discriminatory and invasive forms of data processing.
However, collective interests require adequate forms of representation, as well as the involvement of a range of
stakeholders in the balancing of conflicting interests. Specific procedural criteria must be laid down to define which
entities may act in the collective interest, and this decision is
made more difficult in the context of big data by the lack of
stability in the nature of groups created by data gatherers.
140
In 2014, the 90% of the world’s data were generated in the last
two years, while the remaining 10% were produced through the
rest of humanity history. In 2014, the speed of common computer processors was around 100–200 thousand MIPS (million
instructions per second), in 1994, it was around 180. In 2012, there
were 8.7 billion of connected devices (so-called Internet of Things);
in 2020 they will be 50 billion. See CNN, ‘The data rush: How information about you is 21st century “gold” ’ (November 13, 2014)
<http://edition.cnn.com/2014/11/04/tech/gallery/big-data
-techonomics-graphs/>; Wikipedia, ‘Instructions per second’,
Wikipedia, the free encyclopedia (2015) <https://en.wikipedia.org/wiki/
Instructions_per_second>; Cisco, ‘Seize New IoT Opportunities with
the Cisco IoT System’ <http://www.cisco.com/web/solutions/trends/
iot/portfolio.html>. All these sources were accessed on 12 June 2015.
In this context, the assessment of the social and ethical
impact of analytics may represent an opportunity to discover
how data processing affects collective interests and similarly
an opportunity to identify the potential stakeholders.
Meanwhile, the assessment also represents the principal
mechanism by which conflicting interests relating to the
context-dependent notion of collective data protection can
be balanced.
Finally, given the central role of data processing analysis in
risk-assessment, data protection authorities can play a key role
in the assessment process and licensing models can be reconsidered in the specific context of big data analytics.
255
Acknowledgements
I am indebted to all who provided feedback during the first presentations of my thoughts on this topic at the 6th International
Conference on Information Law and Ethics (Thessaloniki, 30–
31 May 2014) and at the 9th International Conference on Legal,
Security and Privacy Issues in IT Law (Lisbon, 15–17 October
2014). I would like to thank Prof. Lee Bygrave, Prof. Leonardo
Lenti and Dr. Giuseppe Vaciago for their helpful suggestions.
I am grateful to the anonymous CLSR reviewers for their constructive comments.

From an individual to a collective dimension of data protection

Transcript

Documenti analoghi

Strategic Direction Resolution

TECHNICAL NOTE ON HOW TO UPLOAD A YOUTUBE VIDEO On

Legal. - Saab Parts

Library Services Registration

Data Protection, Information Privacy, and Security Measures

Privacy - Aruba Cloud

english - Berliner Beauftragte für Datenschutz und Informationsfreiheit